Abstract. This paper investigates the impacts of educational factors on economic growth across 31 provinces during 1996 and 2010 in China. A spatial panel estimation model is applied to study the impacts of education on economic growth taking into account the spatial spillover effects in Feder model and the cumulative effect. The results reveal that (1) educational factors are significantly spatially autocorrelated. Educational factors have spatial spillover effects. Regional differences of education impacts still exist. (2) Average schooling year has a more positive effect on economic output than capital investment and labor force. Basic education might play a more important role in economic growth. (3) Education sector also benefits non-education sectors on economic growth if "spatial effects of economic shocks" are considered. Some policies that may enhance education development and their impacts on economic growth are proposed.

Recovery energy has been one of important components of energy consumption structure in China. Recovery energy is a type of secondary energy, which is generated from various sources. Since the estimation of recovery energy from some sources such as coke oven gas and other coal gas contains some uncertainties, recovery energy can be represented as an interval variable. The extant studies in the literature ignore recovery energy in the evaluation of energy efficiency or carbon dioxide (CO 2 ) emissions efficiency. As a part of energy supply in China, recovery energy may influence regional primary energy consumption and CO 2 emissions significantly. In this context, this paper proposes an interval slacks-based measure approach to evaluate energy efficiency and CO 2 emissions efficiency in the presence of recovery energy in China. In the described approach, the indicator of CO 2 emissions is incorporated based on the weak disposability assumption, and recovery energy is modeled as a dual factor. The optimal exact data on recovery energy of each region can be obtained. Based on the proposed approach, the measures of energy saving potential and CO 2 emissions reduction potential are derived. Findings resulting from the model application show that there are great disparities in regional efficiencies, and the inefficiency of Chinese regions is largely driven by those inefficient regions in the central and western areas. Notably, the development of recovery energy can help to improve regional energy efficiency and CO 2 emissions efficiency in China.

No abstract